Free AI software
Agents, code editors, local model runners, and tools that work for free — especially when paired with the free APIs in our catalog.
Editor's picks
Self-improving agent that learns skills from experience and keeps cross-session memory; runs in the CLI, a desktop app, Telegram, Discord, Slack and WhatsApp.
Personal AI assistant (formerly Clawdbot/Moltbot) that runs on your own machine and acts through WhatsApp, Telegram, Discord, Slack, iMessage and 20+ channels.
Open-source terminal coding agent with a TUI, desktop app and IDE integration; plan/build agents, LSP support and any model provider.
Google's open-source terminal agent: coding, file and shell tools, web search grounding, MCP servers and a large context window.
Autonomous coding agent for VS Code and JetBrains that plans, edits files, runs commands and uses the browser with your approval.
Microsoft's open-source editor with built-in GitHub Copilot chat, agent mode and inline suggestions; also supports your own model keys.
Google's agent-first development platform: the Antigravity IDE plus an agent manager, CLI and SDK for running coding agents in parallel.
Popular runtime to download and run open LLMs locally via CLI, desktop app and a REST/OpenAI-compatible API.
Desktop app to discover, download and run local LLMs (llama.cpp, MLX) with a chat UI and an OpenAI-compatible local server.
Self-hosted, offline-capable AI chat platform for Ollama and any OpenAI-compatible API, with RAG, tools, pipelines and multi-user.
Fair-code workflow automation with 1,500+ integrations and AI agent/LLM nodes; self-host via Docker or use n8n Cloud.
Node-graph engine for local image, video, 3D and audio generation: SDXL, Flux, Wan 2.x, LTX-2, HunyuanVideo and more.
Offline push-to-talk dictation: hold a hotkey, speak, and the text is pasted into any app; optional LLM cleanup via an OpenAI-compatible API.
AI agents
Agents plan and carry out tasks on their own: writing code and working with files, browsers, and messengers. Everything below is free to run — you only pay for the model, and you can pick one from our free API catalog.
Self-improving agent that learns skills from experience and keeps cross-session memory; runs in the CLI, a desktop app, Telegram, Discord, Slack and WhatsApp.
Personal AI assistant (formerly Clawdbot/Moltbot) that runs on your own machine and acts through WhatsApp, Telegram, Discord, Slack, iMessage and 20+ channels.
Open platform for coding agents: Agent Canvas GUI, CLI and SDK to run OpenHands, Claude Code, Codex or Gemini agents locally, in Docker or in the cloud.
General-purpose agent framework with a Dockerized Linux desktop, browser control, memory, skills, 100+ plugins and multi-agent cooperation.
Python library, CLI and cloud API that let AI agents control a web browser: navigate sites, fill forms and complete multi-step web tasks.
Open-source general AI agent, a Manus alternative without invite codes: tool use, browser automation via Browser Use and MCP support.
Terminal coding agents
Terminal agents read your repository, edit files, and run commands. Most accept any OpenAI-compatible API, so you can connect free models from OpenRouter, NVIDIA Build, or a local Ollama.
Open-source terminal coding agent with a TUI, desktop app and IDE integration; plan/build agents, LSP support and any model provider.
Google's open-source terminal agent: coding, file and shell tools, web search grounding, MCP servers and a large context window.
OpenAI's lightweight coding agent for the terminal, written in Rust, with sandboxed command execution and IDE integrations.
Terminal coding agent adapted from Gemini CLI and optimized for Qwen Coder models.
Stylish terminal coding agent from Charm with LSP context, MCP support and per-session model switching.
AI pair programming in the terminal with a repository map, automatic git commits and support for almost any LLM.
AI code editors
Commercial IDEs cap their free tiers, while open-source editors are fully free and work with your own keys (BYOK) and local models.
Microsoft's open-source editor with built-in GitHub Copilot chat, agent mode and inline suggestions; also supports your own model keys.
AI-first fork of VS Code with agent mode, Tab autocomplete, codebase indexing and multi-model chat.
The Windsurf IDE, renamed by Cognition in June 2026: VS Code-based editor with Devin agents, Tab autocomplete and ACP support.
ByteDance's VS Code-based AI IDE with chat, agent and SOLO modes, autocomplete and MCP support.
Fast open-source Rust editor with an agent panel, edit predictions and ACP support for external agents like Claude Agent and Codex.
AWS agentic IDE built around spec-driven development: prompts become requirements, design and tasks, with hooks and parallel agents.
IDE extensions
Extensions turn your usual editor into an agentic environment. You only pay the model provider — or nothing at all with free routes.
Autonomous coding agent for VS Code and JetBrains that plans, edits files, runs commands and uses the browser with your approval.
Agentic engineering platform for VS Code, JetBrains and the terminal with architect, code and debug modes.
Open-source coding agent with chat, autocomplete and edit modes plus CLI checks that run in CI.
Local LLM runners
Running locally removes quotas and privacy concerns: the model runs on your hardware. Most apps expose an OpenAI-compatible server for IDEs and agents.
Popular runtime to download and run open LLMs locally via CLI, desktop app and a REST/OpenAI-compatible API.
Desktop app to discover, download and run local LLMs (llama.cpp, MLX) with a chat UI and an OpenAI-compatible local server.
Offline ChatGPT alternative: runs local LLMs, connects to cloud APIs, supports MCP and serves an OpenAI-compatible API on localhost.
Reference C/C++ engine for GGUF models; llama-server provides an OpenAI-compatible API and web UI on CPU, CUDA, Metal, Vulkan and ROCm.
High-throughput LLM inference and serving engine with an OpenAI-compatible server for NVIDIA, AMD, Intel, TPU and CPU.
Self-hosted drop-in OpenAI/Anthropic API replacement for LLMs, images, audio, video and embeddings on CPU or GPU.
Chat interfaces
Plug in a free OpenRouter, Groq, or Gemini key and get a full chat with history, RAG, and agents without a subscription.
Self-hosted, offline-capable AI chat platform for Ollama and any OpenAI-compatible API, with RAG, tools, pipelines and multi-user.
Self-hosted multi-provider ChatGPT-style UI with agents, MCP, custom OpenAI-compatible endpoints, RAG and multi-user auth.
Agent-focused AI chat workspace with many providers, Ollama, MCP plugins and a knowledge base; self-host, cloud or desktop.
All-in-one AI app with RAG over documents, agents and MCP; works with local runners and any OpenAI-compatible API.
Desktop multi-LLM client with 300+ assistants, agents and a knowledge base; supports cloud APIs, Ollama/LM Studio and custom endpoints.
Cross-platform desktop chat client for OpenAI, Claude, Gemini, Ollama and other APIs.
Automation and no-code
Visual builders connect LLMs to hundreds of services. Community editions are free when you self-host them.
Fair-code workflow automation with 1,500+ integrations and AI agent/LLM nodes; self-host via Docker or use n8n Cloud.
LLM app platform with visual workflows, RAG, agents and model management; supports Ollama and any OpenAI-compatible API.
Visual low-code builder for AI agents, RAG and MCP flows in Python; each flow can be served as an API or MCP server.
Open-source Zapier alternative with an AI agent builder and 280+ integrations that also work as MCP servers.
Workspace for building AI agents and workflows visually or by chat, with 1,000+ integrations and local models via Ollama/vLLM.
Image and video
Open models such as Flux, Wan, and HunyuanVideo run locally with no per-image cost. You need a GPU; WanGP targets low-VRAM cards.
Node-graph engine for local image, video, 3D and audio generation: SDXL, Flux, Wan 2.x, LTX-2, HunyuanVideo and more.
Canvas-based image studio with layers, inpainting and node workflows; recent versions add Wan 2.2 video and Flux.2.
Actively developed continuation of Forge WebUI (A1111-style) with Flux.2, Qwen-Image, Chroma and Wan 2.2 video support.
Beginner-friendly generation UI on top of ComfyUI for images and video with multi-GPU and multiple backends.
Low-VRAM local app for Wan 2.x, LTX-2, HunyuanVideo, Flux and Qwen Image; some models run on 6 GB GPUs.
One-click launcher and catalog that installs and runs local AI apps (ComfyUI, WanGP, TTS, LLMs) in isolated environments.
Voice and speech
Speech models have long run locally and for free — even on an ordinary laptop.
Offline push-to-talk dictation: hold a hotkey, speak, and the text is pasted into any app; optional LLM cleanup via an OpenAI-compatible API.
OpenAI's multilingual speech recognition and translation models, from tiny to large-v3 and turbo, with a Python CLI.
CTranslate2 reimplementation of Whisper, up to 4x faster with less memory and 8-bit quantization.
Dependency-free C/C++ port of Whisper with CUDA, Vulkan, Metal, quantization, VAD and an HTTP server.
Desktop app to transcribe and translate audio and video offline with Whisper or any OpenAI-compatible API.
Lightweight open-weight 82M-parameter text-to-speech model with 54 voices in 8 languages, fast even on CPU.
FAQ
Which AI tools are completely free?
55 of 73 tools in this catalog are open source: you can run them for free without limits and pay only for the model — or nothing with free APIs and local models.
How do I connect a free model to an agent or IDE?
Get a free key from a provider in our catalog (OpenRouter, Groq, NVIDIA Build, and others), then set its base URL and a model ID in any app that supports the OpenAI-compatible API.
How often is this list updated?
Each entry is checked by hand against the official site or GitHub; the check date is shown on its page. Free tiers change, so confirm on the vendor site before relying on them.